ftp.cc.uoc.gr
Chapter 4. Advanced DNS Features

Chapter 4. Advanced DNS Features

Notify

DNS NOTIFY is a mechanism that allows master servers to notify their slave servers of changes to a zone's data. In response to a NOTIFY from a master server, the slave will check to see that its version of the zone is the current version and, if not, initiate a zone transfer.

For more information about DNS NOTIFY, see the description of the notify option in the section called “Boolean Options” and the description of the zone option also-notify in the section called “Zone Transfers”. The NOTIFY protocol is specified in RFC 1996.

Note

As a slave zone can also be a master to other slaves, named, by default, sends NOTIFY messages for every zone it loads. Specifying notify master-only; will cause named to only send NOTIFY for master zones that it loads.

Dynamic Update

Dynamic Update is a method for adding, replacing or deleting records in a master server by sending it a special form of DNS messages. The format and meaning of these messages is specified in RFC 2136.

Dynamic update is enabled by including an allow-update or an update-policy clause in the zone statement.

If the zone's update-policy is set to local, updates to the zone will be permitted for the key local-ddns, which will be generated by named at startup. See the section called “Dynamic Update Policies” for more details.

The tkey-gssapi-credential and tkey-domain clauses in the options statement enable the server to negotiate keys that can be matched against those in update-policy or allow-update.

Updating of secure zones (zones using DNSSEC) follows RFC 3007: RRSIG, NSEC and NSEC3 records affected by updates are automatically regenerated by the server using an online zone key. Update authorization is based on transaction signatures and an explicit server policy.

The journal file

All changes made to a zone using dynamic update are stored in the zone's journal file. This file is automatically created by the server when the first dynamic update takes place. The name of the journal file is formed by appending the extension .jnl to the name of the corresponding zone file unless specifically overridden. The journal file is in a binary format and should not be edited manually.

The server will also occasionally write ("dump") the complete contents of the updated zone to its zone file. This is not done immediately after each dynamic update, because that would be too slow when a large zone is updated frequently. Instead, the dump is delayed by up to 15 minutes, allowing additional updates to take place. During the dump process, transient files will be created with the extensions .jnw and .jbk; under ordinary circumstances, these will be removed when the dump is complete, and can be safely ignored.

When a server is restarted after a shutdown or crash, it will replay the journal file to incorporate into the zone any updates that took place after the last zone dump.

Changes that result from incoming incremental zone transfers are also journalled in a similar way.

The zone files of dynamic zones cannot normally be edited by hand because they are not guaranteed to contain the most recent dynamic changes — those are only in the journal file. The only way to ensure that the zone file of a dynamic zone is up to date is to run rndc stop.

If you have to make changes to a dynamic zone manually, the following procedure will work: Disable dynamic updates to the zone using rndc freeze zone. This will also remove the zone's .jnl file and update the master file. Edit the zone file. Run rndc thaw zone to reload the changed zone and re-enable dynamic updates.

Incremental Zone Transfers (IXFR)

The incremental zone transfer (IXFR) protocol is a way for slave servers to transfer only changed data, instead of having to transfer the entire zone. The IXFR protocol is specified in RFC 1995. See Proposed Standards.

When acting as a master, BIND 9 supports IXFR for those zones where the necessary change history information is available. These include master zones maintained by dynamic update and slave zones whose data was obtained by IXFR. For manually maintained master zones, and for slave zones obtained by performing a full zone transfer (AXFR), IXFR is supported only if the option ixfr-from-differences is set to yes.

When acting as a slave, BIND 9 will attempt to use IXFR unless it is explicitly disabled via the request-ixfr option or the use of ixfr-from-differences. For more information about disabling IXFR, see the description of the request-ixfr clause of the server statement.

Split DNS

Setting up different views, or visibility, of the DNS space to internal and external resolvers is usually referred to as a Split DNS setup. There are several reasons an organization would want to set up its DNS this way.

One common reason for setting up a DNS system this way is to hide "internal" DNS information from "external" clients on the Internet. There is some debate as to whether or not this is actually useful. Internal DNS information leaks out in many ways (via email headers, for example) and most savvy "attackers" can find the information they need using other means. However, since listing addresses of internal servers that external clients cannot possibly reach can result in connection delays and other annoyances, an organization may choose to use a Split DNS to present a consistent view of itself to the outside world.

Another common reason for setting up a Split DNS system is to allow internal networks that are behind filters or in RFC 1918 space (reserved IP space, as documented in RFC 1918) to resolve DNS on the Internet. Split DNS can also be used to allow mail from outside back in to the internal network.

Example split DNS setup

Let's say a company named Example, Inc. (example.com) has several corporate sites that have an internal network with reserved Internet Protocol (IP) space and an external demilitarized zone (DMZ), or "outside" section of a network, that is available to the public.

Example, Inc. wants its internal clients to be able to resolve external hostnames and to exchange mail with people on the outside. The company also wants its internal resolvers to have access to certain internal-only zones that are not available at all outside of the internal network.

In order to accomplish this, the company will set up two sets of name servers. One set will be on the inside network (in the reserved IP space) and the other set will be on bastion hosts, which are "proxy" hosts that can talk to both sides of its network, in the DMZ.

The internal servers will be configured to forward all queries, except queries for site1.internal, site2.internal, site1.example.com, and site2.example.com, to the servers in the DMZ. These internal servers will have complete sets of information for site1.example.com, site2.example.com, site1.internal, and site2.internal.

To protect the site1.internal and site2.internal domains, the internal name servers must be configured to disallow all queries to these domains from any external hosts, including the bastion hosts.

The external servers, which are on the bastion hosts, will be configured to serve the "public" version of the site1 and site2.example.com zones. This could include things such as the host records for public servers (www.example.com and ftp.example.com), and mail exchange (MX) records (a.mx.example.com and b.mx.example.com).

In addition, the public site1 and site2.example.com zones should have special MX records that contain wildcard (`*') records pointing to the bastion hosts. This is needed because external mail servers do not have any other way of looking up how to deliver mail to those internal hosts. With the wildcard records, the mail will be delivered to the bastion host, which can then forward it on to internal hosts.

Here's an example of a wildcard MX record:

*   IN MX 10 external1.example.com.

Now that they accept mail on behalf of anything in the internal network, the bastion hosts will need to know how to deliver mail to internal hosts. In order for this to work properly, the resolvers on the bastion hosts will need to be configured to point to the internal name servers for DNS resolution.

Queries for internal hostnames will be answered by the internal servers, and queries for external hostnames will be forwarded back out to the DNS servers on the bastion hosts.

In order for all this to work properly, internal clients will need to be configured to query only the internal name servers for DNS queries. This could also be enforced via selective filtering on the network.

If everything has been set properly, Example, Inc.'s internal clients will now be able to:

  • Look up any hostnames in the site1 and site2.example.com zones.
  • Look up any hostnames in the site1.internal and site2.internal domains.
  • Look up any hostnames on the Internet.
  • Exchange mail with both internal and external people.

Hosts on the Internet will be able to:

  • Look up any hostnames in the site1 and site2.example.com zones.
  • Exchange mail with anyone in the site1 and site2.example.com zones.

Here is an example configuration for the setup we just described above. Note that this is only configuration information; for information on how to configure your zone files, see the section called “Sample Configurations”.

Internal DNS server config:


acl internals { 172.16.72.0/24; 192.168.1.0/24; };

acl externals { bastion-ips-go-here; };

options {
    ...
    ...
    forward only;
    // forward to external servers
    forwarders {
        bastion-ips-go-here;
    };
    // sample allow-transfer (no one)
    allow-transfer { none; };
    // restrict query access
    allow-query { internals; externals; };
    // restrict recursion
    allow-recursion { internals; };
    ...
    ...
};

// sample master zone
zone "site1.example.com" {
  type master;
  file "m/site1.example.com";
  // do normal iterative resolution (do not forward)
  forwarders { };
  allow-query { internals; externals; };
  allow-transfer { internals; };
};

// sample slave zone
zone "site2.example.com" {
  type slave;
  file "s/site2.example.com";
  masters { 172.16.72.3; };
  forwarders { };
  allow-query { internals; externals; };
  allow-transfer { internals; };
};

zone "site1.internal" {
  type master;
  file "m/site1.internal";
  forwarders { };
  allow-query { internals; };
  allow-transfer { internals; }
};

zone "site2.internal" {
  type slave;
  file "s/site2.internal";
  masters { 172.16.72.3; };
  forwarders { };
  allow-query { internals };
  allow-transfer { internals; }
};

External (bastion host) DNS server config:

acl internals { 172.16.72.0/24; 192.168.1.0/24; };

acl externals { bastion-ips-go-here; };

options {
  ...
  ...
  // sample allow-transfer (no one)
  allow-transfer { none; };
  // default query access
  allow-query { any; };
  // restrict cache access
  allow-query-cache { internals; externals; };
  // restrict recursion
  allow-recursion { internals; externals; };
  ...
  ...
};

// sample slave zone
zone "site1.example.com" {
  type master;
  file "m/site1.foo.com";
  allow-transfer { internals; externals; };
};

zone "site2.example.com" {
  type slave;
  file "s/site2.foo.com";
  masters { another_bastion_host_maybe; };
  allow-transfer { internals; externals; }
};

In the resolv.conf (or equivalent) on the bastion host(s):

search ...
nameserver 172.16.72.2
nameserver 172.16.72.3
nameserver 172.16.72.4

TSIG

This is a short guide to setting up Transaction SIGnatures (TSIG) based transaction security in BIND. It describes changes to the configuration file as well as what changes are required for different features, including the process of creating transaction keys and using transaction signatures with BIND.

BIND primarily supports TSIG for server to server communication. This includes zone transfer, notify, and recursive query messages. Resolvers based on newer versions of BIND 8 have limited support for TSIG.

TSIG can also be useful for dynamic update. A primary server for a dynamic zone should control access to the dynamic update service, but IP-based access control is insufficient. The cryptographic access control provided by TSIG is far superior. The nsupdate program supports TSIG via the -k and -y command line options or inline by use of the key.

Generate Shared Keys for Each Pair of Hosts

A shared secret is generated to be shared between host1 and host2. An arbitrary key name is chosen: "host1-host2.". The key name must be the same on both hosts.

Automatic Generation

The following command will generate a 128-bit (16 byte) HMAC-SHA256 key as described above. Longer keys are better, but shorter keys are easier to read. Note that the maximum key length is the digest length, here 256 bits.

dnssec-keygen -a hmac-sha256 -b 128 -n HOST host1-host2.

The key is in the file Khost1-host2.+163+00000.private. Nothing directly uses this file, but the base-64 encoded string following "Key:" can be extracted from the file and used as a shared secret:

Key: La/E5CjG9O+os1jq0a2jdA==

The string "La/E5CjG9O+os1jq0a2jdA==" can be used as the shared secret.

Manual Generation

The shared secret is simply a random sequence of bits, encoded in base-64. Most ASCII strings are valid base-64 strings (assuming the length is a multiple of 4 and only valid characters are used), so the shared secret can be manually generated.

Also, a known string can be run through mmencode or a similar program to generate base-64 encoded data.

Copying the Shared Secret to Both Machines

This is beyond the scope of DNS. A secure transport mechanism should be used. This could be secure FTP, ssh, telephone, etc.

Informing the Servers of the Key's Existence

Imagine host1 and host 2 are both servers. The following is added to each server's named.conf file:

key host1-host2. {
  algorithm hmac-sha256;
  secret "La/E5CjG9O+os1jq0a2jdA==";
};

The secret is the one generated above. Since this is a secret, it is recommended that either named.conf be non-world readable, or the key directive be added to a non-world readable file that is included by named.conf.

At this point, the key is recognized. This means that if the server receives a message signed by this key, it can verify the signature. If the signature is successfully verified, the response is signed by the same key.

Instructing the Server to Use the Key

Since keys are shared between two hosts only, the server must be told when keys are to be used. The following is added to the named.conf file for host1, if the IP address of host2 is 10.1.2.3:

server 10.1.2.3 {
  keys { host1-host2. ;};
};

Multiple keys may be present, but only the first is used. This directive does not contain any secrets, so it may be in a world-readable file.

If host1 sends a message that is a request to that address, the message will be signed with the specified key. host1 will expect any responses to signed messages to be signed with the same key.

A similar statement must be present in host2's configuration file (with host1's address) for host2 to sign request messages to host1.

TSIG Key Based Access Control

BIND allows IP addresses and ranges to be specified in ACL definitions and allow-{ query | transfer | update } directives. This has been extended to allow TSIG keys also. The above key would be denoted key host1-host2.

An example of an allow-update directive would be:

allow-update { key host1-host2. ;};

This allows dynamic updates to succeed only if the request was signed by a key named "host1-host2.".

See the section called “Dynamic Update Policies” for a discussion of the more flexible update-policy statement.

Errors

The processing of TSIG signed messages can result in several errors. If a signed message is sent to a non-TSIG aware server, a FORMERR (format error) will be returned, since the server will not understand the record. This is a result of misconfiguration, since the server must be explicitly configured to send a TSIG signed message to a specific server.

If a TSIG aware server receives a message signed by an unknown key, the response will be unsigned with the TSIG extended error code set to BADKEY. If a TSIG aware server receives a message with a signature that does not validate, the response will be unsigned with the TSIG extended error code set to BADSIG. If a TSIG aware server receives a message with a time outside of the allowed range, the response will be signed with the TSIG extended error code set to BADTIME, and the time values will be adjusted so that the response can be successfully verified. In any of these cases, the message's rcode (response code) is set to NOTAUTH (not authenticated).

TKEY

TKEY is a mechanism for automatically generating a shared secret between two hosts. There are several "modes" of TKEY that specify how the key is generated or assigned. BIND 9 implements only one of these modes, the Diffie-Hellman key exchange. Both hosts are required to have a Diffie-Hellman KEY record (although this record is not required to be present in a zone). The TKEY process must use signed messages, signed either by TSIG or SIG(0). The result of TKEY is a shared secret that can be used to sign messages with TSIG. TKEY can also be used to delete shared secrets that it had previously generated.

The TKEY process is initiated by a client or server by sending a signed TKEY query (including any appropriate KEYs) to a TKEY-aware server. The server response, if it indicates success, will contain a TKEY record and any appropriate keys. After this exchange, both participants have enough information to determine the shared secret; the exact process depends on the TKEY mode. When using the Diffie-Hellman TKEY mode, Diffie-Hellman keys are exchanged, and the shared secret is derived by both participants.

SIG(0)

BIND 9 partially supports DNSSEC SIG(0) transaction signatures as specified in RFC 2535 and RFC 2931. SIG(0) uses public/private keys to authenticate messages. Access control is performed in the same manner as TSIG keys; privileges can be granted or denied based on the key name.

When a SIG(0) signed message is received, it will only be verified if the key is known and trusted by the server; the server will not attempt to locate and/or validate the key.

SIG(0) signing of multiple-message TCP streams is not supported.

The only tool shipped with BIND 9 that generates SIG(0) signed messages is nsupdate.

DNSSEC

Cryptographic authentication of DNS information is possible through the DNS Security (DNSSEC-bis) extensions, defined in RFC 4033, RFC 4034, and RFC 4035. This section describes the creation and use of DNSSEC signed zones.

In order to set up a DNSSEC secure zone, there are a series of steps which must be followed. BIND 9 ships with several tools that are used in this process, which are explained in more detail below. In all cases, the -h option prints a full list of parameters. Note that the DNSSEC tools require the keyset files to be in the working directory or the directory specified by the -d option, and that the tools shipped with BIND 9.2.x and earlier are not compatible with the current ones.

There must also be communication with the administrators of the parent and/or child zone to transmit keys. A zone's security status must be indicated by the parent zone for a DNSSEC capable resolver to trust its data. This is done through the presence or absence of a DS record at the delegation point.

For other servers to trust data in this zone, they must either be statically configured with this zone's zone key or the zone key of another zone above this one in the DNS tree.

Generating Keys

The dnssec-keygen program is used to generate keys.

A secure zone must contain one or more zone keys. The zone keys will sign all other records in the zone, as well as the zone keys of any secure delegated zones. Zone keys must have the same name as the zone, a name type of ZONE, and must be usable for authentication. It is recommended that zone keys use a cryptographic algorithm designated as "mandatory to implement" by the IETF; currently the only one is RSASHA1.

The following command will generate a 768-bit RSASHA1 key for the child.example zone:

dnssec-keygen -a RSASHA1 -b 768 -n ZONE child.example.

Two output files will be produced: Kchild.example.+005+12345.key and Kchild.example.+005+12345.private (where 12345 is an example of a key tag). The key filenames contain the key name (child.example.), algorithm (3 is DSA, 1 is RSAMD5, 5 is RSASHA1, etc.), and the key tag (12345 in this case). The private key (in the .private file) is used to generate signatures, and the public key (in the .key file) is used for signature verification.

To generate another key with the same properties (but with a different key tag), repeat the above command.

The dnssec-keyfromlabel program is used to get a key pair from a crypto hardware and build the key files. Its usage is similar to dnssec-keygen.

The public keys should be inserted into the zone file by including the .key files using $INCLUDE statements.

Signing the Zone

The dnssec-signzone program is used to sign a zone.

Any keyset files corresponding to secure subzones should be present. The zone signer will generate NSEC, NSEC3 and RRSIG records for the zone, as well as DS for the child zones if '-g' is specified. If '-g' is not specified, then DS RRsets for the secure child zones need to be added manually.

The following command signs the zone, assuming it is in a file called zone.child.example. By default, all zone keys which have an available private key are used to generate signatures.

dnssec-signzone -o child.example zone.child.example

One output file is produced: zone.child.example.signed. This file should be referenced by named.conf as the input file for the zone.

dnssec-signzone will also produce a keyset and dsset files and optionally a dlvset file. These are used to provide the parent zone administrators with the DNSKEYs (or their corresponding DS records) that are the secure entry point to the zone.

Configuring Servers

To enable named to respond appropriately to DNS requests from DNSSEC aware clients, dnssec-enable must be set to yes. (This is the default setting.)

To enable named to validate answers from other servers, the dnssec-enable and dnssec-validation options must both be set to yes (the default setting in BIND 9.5 and later), and at least one trust anchor must be configured with a trusted-keys or managed-keys statement in named.conf.

trusted-keys are copies of DNSKEY RRs for zones that are used to form the first link in the cryptographic chain of trust. All keys listed in trusted-keys (and corresponding zones) are deemed to exist and only the listed keys will be used to validated the DNSKEY RRset that they are from.

managed-keys are trusted keys which are automatically kept up to date via RFC 5011 trust anchor maintenance.

trusted-keys and managed-keys are described in more detail later in this document.

Unlike BIND 8, BIND 9 does not verify signatures on load, so zone keys for authoritative zones do not need to be specified in the configuration file.

After DNSSEC gets established, a typical DNSSEC configuration will look something like the following. It has one or more public keys for the root. This allows answers from outside the organization to be validated. It will also have several keys for parts of the namespace the organization controls. These are here to ensure that named is immune to compromises in the DNSSEC components of the security of parent zones.

managed-keys {
        /* Root Key */
        "." initial-key 257 3 3 "BNY4wrWM1nCfJ+CXd0rVXyYmobt7sEEfK3clRbGaTwS
                                 JxrGkxJWoZu6I7PzJu/E9gx4UC1zGAHlXKdE4zYIpRh
                                 aBKnvcC2U9mZhkdUpd1Vso/HAdjNe8LmMlnzY3zy2Xy
                                 4klWOADTPzSv9eamj8V18PHGjBLaVtYvk/ln5ZApjYg
                                 hf+6fElrmLkdaz MQ2OCnACR817DF4BBa7UR/beDHyp
                                 5iWTXWSi6XmoJLbG9Scqc7l70KDqlvXR3M/lUUVRbke
                                 g1IPJSidmK3ZyCllh4XSKbje/45SKucHgnwU5jefMtq
                                 66gKodQj+MiA21AfUVe7u99WzTLzY3qlxDhxYQQ20FQ
                                 97S+LKUTpQcq27R7AT3/V5hRQxScINqwcz4jYqZD2fQ
                                 dgxbcDTClU0CRBdiieyLMNzXG3";
};

trusted-keys {
        /* Key for our organization's forward zone */
        example.com. 257 3 5 "AwEAAaxPMcR2x0HbQV4WeZB6oEDX+r0QM6
                              5KbhTjrW1ZaARmPhEZZe3Y9ifgEuq7vZ/z
                              GZUdEGNWy+JZzus0lUptwgjGwhUS1558Hb
                              4JKUbbOTcM8pwXlj0EiX3oDFVmjHO444gL
                              kBOUKUf/mC7HvfwYH/Be22GnClrinKJp1O
                              g4ywzO9WglMk7jbfW33gUKvirTHr25GL7S
                              TQUzBb5Usxt8lgnyTUHs1t3JwCY5hKZ6Cq
                              FxmAVZP20igTixin/1LcrgX/KMEGd/biuv
                              F4qJCyduieHukuY3H4XMAcR+xia2nIUPvm
                              /oyWR8BW/hWdzOvnSCThlHf3xiYleDbt/o
                              1OTQ09A0=";

        /* Key for our reverse zone. */
        2.0.192.IN-ADDRPA.NET. 257 3 5 "AQOnS4xn/IgOUpBPJ3bogzwc
                                       xOdNax071L18QqZnQQQAVVr+i
                                       LhGTnNGp3HoWQLUIzKrJVZ3zg
                                       gy3WwNT6kZo6c0tszYqbtvchm
                                       gQC8CzKojM/W16i6MG/eafGU3
                                       siaOdS0yOI6BgPsw+YZdzlYMa
                                       IJGf4M4dyoKIhzdZyQ2bYQrjy
                                       Q4LB0lC7aOnsMyYKHHYeRvPxj
                                       IQXmdqgOJGq+vsevG06zW+1xg
                                       YJh9rCIfnm1GX/KMgxLPG2vXT
                                       D/RnLX+D3T3UL7HJYHJhAZD5L
                                       59VvjSPsZJHeDCUyWYrvPZesZ
                                       DIRvhDD52SKvbheeTJUm6Ehkz
                                       ytNN2SN96QRk8j/iI8ib";
};

options {
        ...
        dnssec-enable yes;
        dnssec-validation yes;
};

Note

None of the keys listed in this example are valid. In particular, the root key is not valid.

When DNSSEC validation is enabled and properly configured, the resolver will reject any answers from signed, secure zones which fail to validate, and will return SERVFAIL to the client.

Responses may fail to validate for any of several reasons, including missing, expired, or invalid signatures, a key which does not match the DS RRset in the parent zone, or an insecure response from a zone which, according to its parent, should have been secure.

Note

When the validator receives a response from an unsigned zone that has a signed parent, it must confirm with the parent that the zone was intentionally left unsigned. It does this by verifying, via signed and validated NSEC/NSEC3 records, that the parent zone contains no DS records for the child.

If the validator can prove that the zone is insecure, then the response is accepted. However, if it cannot, then it must assume an insecure response to be a forgery; it rejects the response and logs an error.

The logged error reads "insecurity proof failed" and "got insecure response; parent indicates it should be secure". (Prior to BIND 9.7, the logged error was "not insecure". This referred to the zone, not the response.)

DNSSEC, Dynamic Zones, and Automatic Signing

As of BIND 9.7.0 it is possible to change a dynamic zone from insecure to signed and back again. A secure zone can use either NSEC or NSEC3 chains.

Converting from insecure to secure

Changing a zone from insecure to secure can be done in two ways: using a dynamic DNS update, or the auto-dnssec zone option.

For either method, you need to configure named so that it can see the K* files which contain the public and private parts of the keys that will be used to sign the zone. These files will have been generated by dnssec-keygen. You can do this by placing them in the key-directory, as specified in named.conf:

        zone example.net {
                type master;
                update-policy local;
                file "dynamic/example.net/example.net";
                key-directory "dynamic/example.net";
        };

If one KSK and one ZSK DNSKEY key have been generated, this configuration will cause all records in the zone to be signed with the ZSK, and the DNSKEY RRset to be signed with the KSK as well. An NSEC chain will be generated as part of the initial signing process.

Dynamic DNS update method

To insert the keys via dynamic update:

        % nsupdate
        > ttl 3600
        > update add example.net DNSKEY 256 3 7 AwEAAZn17pUF0KpbPA2c7Gz76Vb18v0teKT3EyAGfBfL8eQ8al35zz3Y I1m/SAQBxIqMfLtIwqWPdgthsu36azGQAX8=
        > update add example.net DNSKEY 257 3 7 AwEAAd/7odU/64o2LGsifbLtQmtO8dFDtTAZXSX2+X3e/UNlq9IHq3Y0 XtC0Iuawl/qkaKVxXe2lo8Ct+dM6UehyCqk=
        > send

While the update request will complete almost immediately, the zone will not be completely signed until named has had time to walk the zone and generate the NSEC and RRSIG records. The NSEC record at the apex will be added last, to signal that there is a complete NSEC chain.

If you wish to sign using NSEC3 instead of NSEC, you should add an NSEC3PARAM record to the initial update request. If you wish the NSEC3 chain to have the OPTOUT bit set, set it in the flags field of the NSEC3PARAM record.

        % nsupdate
        > ttl 3600
        > update add example.net DNSKEY 256 3 7 AwEAAZn17pUF0KpbPA2c7Gz76Vb18v0teKT3EyAGfBfL8eQ8al35zz3Y I1m/SAQBxIqMfLtIwqWPdgthsu36azGQAX8=
        > update add example.net DNSKEY 257 3 7 AwEAAd/7odU/64o2LGsifbLtQmtO8dFDtTAZXSX2+X3e/UNlq9IHq3Y0 XtC0Iuawl/qkaKVxXe2lo8Ct+dM6UehyCqk=
        > update add example.net NSEC3PARAM 1 1 100 1234567890
        > send

Again, this update request will complete almost immediately; however, the record won't show up until named has had a chance to build/remove the relevant chain. A private type record will be created to record the state of the operation (see below for more details), and will be removed once the operation completes.

While the initial signing and NSEC/NSEC3 chain generation is happening, other updates are possible as well.

Fully automatic zone signing

To enable automatic signing, add the auto-dnssec option to the zone statement in named.conf. auto-dnssec has two possible arguments: allow or maintain.

With auto-dnssec allow, named can search the key directory for keys matching the zone, insert them into the zone, and use them to sign the zone. It will do so only when it receives an rndc sign <zonename> or rndc loadkeys <zonename> command.

auto-dnssec maintain includes the above functionality, but will also automatically adjust the zone's DNSKEY records on schedule according to the keys' timing metadata. (See dnssec-keygen(8) and dnssec-settime(8) for more information.) If keys are present in the key directory the first time the zone is loaded, it will be signed immediately, without waiting for an rndc sign or rndc loadkeys command. (Those commands can still be used when there are unscheduled key changes, however.)

Using the auto-dnssec option requires the zone to be configured to allow dynamic updates, by adding an allow-update or update-policy statement to the zone configuration. If this has not been done, the configuration will fail.

Private-type records

The state of the signing process is signaled by private-type records (with a default type value of 65534). When signing is complete, these records will have a nonzero value for the final octet (for those records which have a nonzero initial octet).

The private type record format: If the first octet is non-zero then the record indicates that the zone needs to be signed with the key matching the record, or that all signatures that match the record should be removed.



  algorithm (octet 1)
  key id in network order (octet 2 and 3)
  removal flag (octet 4)
  complete flag (octet 5)

Only records flagged as "complete" can be removed via dynamic update. Attempts to remove other private type records will be silently ignored.

If the first octet is zero (this is a reserved algorithm number that should never appear in a DNSKEY record) then the record indicates changes to the NSEC3 chains are in progress. The rest of the record contains an NSEC3PARAM record. The flag field tells what operation to perform based on the flag bits.



  0x01 OPTOUT
  0x80 CREATE
  0x40 REMOVE
  0x20 NONSEC

DNSKEY rollovers

As with insecure-to-secure conversions, rolling DNSSEC keys can be done in two ways: using a dynamic DNS update, or the auto-dnssec zone option.

Dynamic DNS update method

To perform key rollovers via dynamic update, you need to add the K* files for the new keys so that named can find them. You can then add the new DNSKEY RRs via dynamic update. named will then cause the zone to be signed with the new keys. When the signing is complete the private type records will be updated so that the last octet is non zero.

If this is for a KSK you need to inform the parent and any trust anchor repositories of the new KSK.

You should then wait for the maximum TTL in the zone before removing the old DNSKEY. If it is a KSK that is being updated, you also need to wait for the DS RRset in the parent to be updated and its TTL to expire. This ensures that all clients will be able to verify at least one signature when you remove the old DNSKEY.

The old DNSKEY can be removed via UPDATE. Take care to specify the correct key. named will clean out any signatures generated by the old key after the update completes.

Automatic key rollovers

When a new key reaches its activation date (as set by dnssec-keygen or dnssec-settime), if the auto-dnssec zone option is set to maintain, named will automatically carry out the key rollover. If the key's algorithm has not previously been used to sign the zone, then the zone will be fully signed as quickly as possible. However, if the new key is replacing an existing key of the same algorithm, then the zone will be re-signed incrementally, with signatures from the old key being replaced with signatures from the new key as their signature validity periods expire. By default, this rollover completes in 30 days, after which it will be safe to remove the old key from the DNSKEY RRset.

NSEC3PARAM rollovers via UPDATE

Add the new NSEC3PARAM record via dynamic update. When the new NSEC3 chain has been generated, the NSEC3PARAM flag field will be zero. At this point you can remove the old NSEC3PARAM record. The old chain will be removed after the update request completes.

Converting from NSEC to NSEC3

To do this, you just need to add an NSEC3PARAM record. When the conversion is complete, the NSEC chain will have been removed and the NSEC3PARAM record will have a zero flag field. The NSEC3 chain will be generated before the NSEC chain is destroyed.

Converting from NSEC3 to NSEC

To do this, use nsupdate to remove all NSEC3PARAM records with a zero flag field. The NSEC chain will be generated before the NSEC3 chain is removed.

Converting from secure to insecure

To convert a signed zone to unsigned using dynamic DNS, delete all the DNSKEY records from the zone apex using nsupdate. All signatures, NSEC or NSEC3 chains, and associated NSEC3PARAM records will be removed automatically. This will take place after the update request completes.

This requires the dnssec-secure-to-insecure option to be set to yes in named.conf.

In addition, if the auto-dnssec maintain zone statement is used, it should be removed or changed to allow instead (or it will re-sign).

Periodic re-signing

In any secure zone which supports dynamic updates, named will periodically re-sign RRsets which have not been re-signed as a result of some update action. The signature lifetimes will be adjusted so as to spread the re-sign load over time rather than all at once.

NSEC3 and OPTOUT

named only supports creating new NSEC3 chains where all the NSEC3 records in the zone have the same OPTOUT state. named supports UPDATES to zones where the NSEC3 records in the chain have mixed OPTOUT state. named does not support changing the OPTOUT state of an individual NSEC3 record, the entire chain needs to be changed if the OPTOUT state of an individual NSEC3 needs to be changed.

Dynamic Trust Anchor Management

BIND 9.7.0 introduces support for RFC 5011, dynamic trust anchor management. Using this feature allows named to keep track of changes to critical DNSSEC keys without any need for the operator to make changes to configuration files.

Validating Resolver

To configure a validating resolver to use RFC 5011 to maintain a trust anchor, configure the trust anchor using a managed-keys statement. Information about this can be found in the section called “managed-keys Statement Definition and Usage”.

Authoritative Server

To set up an authoritative zone for RFC 5011 trust anchor maintenance, generate two (or more) key signing keys (KSKs) for the zone. Sign the zone with one of them; this is the "active" KSK. All KSK's which do not sign the zone are "stand-by" keys.

Any validating resolver which is configured to use the active KSK as an RFC 5011-managed trust anchor will take note of the stand-by KSKs in the zone's DNSKEY RRset, and store them for future reference. The resolver will recheck the zone periodically, and after 30 days, if the new key is still there, then the key will be accepted by the resolver as a valid trust anchor for the zone. Any time after this 30-day acceptance timer has completed, the active KSK can be revoked, and the zone can be "rolled over" to the newly accepted key.

The easiest way to place a stand-by key in a zone is to use the "smart signing" features of dnssec-keygen and dnssec-signzone. If a key with a publication date in the past, but an activation date which is unset or in the future, " dnssec-signzone -S" will include the DNSKEY record in the zone, but will not sign with it:

$ dnssec-keygen -K keys -f KSK -P now -A now+2y example.net
$ dnssec-signzone -S -K keys example.net

To revoke a key, the new command dnssec-revoke has been added. This adds the REVOKED bit to the key flags and re-generates the K*.key and K*.private files.

After revoking the active key, the zone must be signed with both the revoked KSK and the new active KSK. (Smart signing takes care of this automatically.)

Once a key has been revoked and used to sign the DNSKEY RRset in which it appears, that key will never again be accepted as a valid trust anchor by the resolver. However, validation can proceed using the new active key (which had been accepted by the resolver when it was a stand-by key).

See RFC 5011 for more details on key rollover scenarios.

When a key has been revoked, its key ID changes, increasing by 128, and wrapping around at 65535. So, for example, the key "Kexample.com.+005+10000" becomes "Kexample.com.+005+10128".

If two keys have ID's exactly 128 apart, and one is revoked, then the two key ID's will collide, causing several problems. To prevent this, dnssec-keygen will not generate a new key if another key is present which may collide. This checking will only occur if the new keys are written to the same directory which holds all other keys in use for that zone.

Older versions of BIND 9 did not have this precaution. Exercise caution if using key revocation on keys that were generated by previous releases, or if using keys stored in multiple directories or on multiple machines.

It is expected that a future release of BIND 9 will address this problem in a different way, by storing revoked keys with their original unrevoked key ID's.

PKCS #11 (Cryptoki) support

PKCS #11 (Public Key Cryptography Standard #11) defines a platform- independent API for the control of hardware security modules (HSMs) and other cryptographic support devices.

BIND 9 is known to work with two HSMs: The Sun SCA 6000 cryptographic acceleration board, tested under Solaris x86, and the AEP Keyper network-attached key storage device, tested with Debian Linux, Solaris x86 and Windows Server 2003.

Prerequisites

See the HSM vendor documentation for information about installing, initializing, testing and troubleshooting the HSM.

BIND 9 uses OpenSSL for cryptography, but stock OpenSSL does not yet fully support PKCS #11. However, a PKCS #11 engine for OpenSSL is available from the OpenSolaris project. It has been modified by ISC to work with with BIND 9, and to provide new features such as PIN management and key by reference.

The patched OpenSSL depends on a "PKCS #11 provider". This is a shared library object, providing a low-level PKCS #11 interface to the HSM hardware. It is dynamically loaded by OpenSSL at runtime. The PKCS #11 provider comes from the HSM vendor, and and is specific to the HSM to be controlled.

There are two "flavors" of PKCS #11 support provided by the patched OpenSSL, one of which must be chosen at configuration time. The correct choice depends on the HSM hardware:

  • Use 'crypto-accelerator' with HSMs that have hardware cryptographic acceleration features, such as the SCA 6000 board. This causes OpenSSL to run all supported cryptographic operations in the HSM.

  • Use 'sign-only' with HSMs that are designed to function primarily as secure key storage devices, but lack hardware acceleration. These devices are highly secure, but are not necessarily any faster at cryptography than the system CPU — often, they are slower. It is therefore most efficient to use them only for those cryptographic functions that require access to the secured private key, such as zone signing, and to use the system CPU for all other computationally-intensive operations. The AEP Keyper is an example of such a device.

The modified OpenSSL code is included in the BIND 9 release, in the form of a context diff against the latest verions of OpenSSL. OpenSSL 0.9.8 and 1.0.0 are both supported; there are separate diffs for each version. In the examples to follow, we use OpenSSL 0.9.8, but the same methods work with OpenSSL 1.0.0.

Note

The latest OpenSSL versions at the time of the BIND release are 0.9.8s and 1.0.0f. ISC will provide an updated patch as new versions of OpenSSL are released. The version number in the following examples is expected to change.

Before building BIND 9 with PKCS #11 support, it will be necessary to build OpenSSL with this patch in place and inform it of the path to the HSM-specific PKCS #11 provider library.

Obtain OpenSSL 0.9.8s:

$ wget http://www.openssl.org/source/openssl-0.9.8s.tar.gz

Extract the tarball:

$ tar zxf openssl-0.9.8s.tar.gz

Apply the patch from the BIND 9 release:

$ patch -p1 -d openssl-0.9.8s \
            < bind9/bin/pkcs11/openssl-0.9.8s-patch

Note

(Note that the patch file may not be compatible with the "patch" utility on all operating systems. You may need to install GNU patch.)

When building OpenSSL, place it in a non-standard location so that it does not interfere with OpenSSL libraries elsewhere on the system. In the following examples, we choose to install into "/opt/pkcs11/usr". We will use this location when we configure BIND 9.

Building OpenSSL for the AEP Keyper on Linux

The AEP Keyper is a highly secure key storage device, but does not provide hardware cryptographic acceleration. It can carry out cryptographic operations, but it is probably slower than your system's CPU. Therefore, we choose the 'sign-only' flavor when building OpenSSL.

The Keyper-specific PKCS #11 provider library is delivered with the Keyper software. In this example, we place it /opt/pkcs11/usr/lib:

$ cp pkcs11.GCC4.0.2.so.4.05 /opt/pkcs11/usr/lib/libpkcs11.so

This library is only available for Linux as a 32-bit binary. If we are compiling on a 64-bit Linux system, it is necessary to force a 32-bit build, by specifying -m32 in the build options.

Finally, the Keyper library requires threads, so we must specify -pthread.

$ cd openssl-0.9.8s
$ ./Configure linux-generic32 -m32 -pthread \
            --pk11-libname=/opt/pkcs11/usr/lib/libpkcs11.so \
            --pk11-flavor=sign-only \
            --prefix=/opt/pkcs11/usr

After configuring, run "make" and "make test". If "make test" fails with "pthread_atfork() not found", you forgot to add the -pthread above.

Building OpenSSL for the SCA 6000 on Solaris

The SCA-6000 PKCS #11 provider is installed as a system library, libpkcs11. It is a true crypto accelerator, up to 4 times faster than any CPU, so the flavor shall be 'crypto-accelerator'.

In this example, we are building on Solaris x86 on an AMD64 system.

$ cd openssl-0.9.8s
$ ./Configure solaris64-x86_64-cc \
            --pk11-libname=/usr/lib/64/libpkcs11.so \
            --pk11-flavor=crypto-accelerator \
            --prefix=/opt/pkcs11/usr

(For a 32-bit build, use "solaris-x86-cc" and /usr/lib/libpkcs11.so.)

After configuring, run make and make test.

Building OpenSSL for SoftHSM

SoftHSM is a software library provided by the OpenDNSSEC project (http://www.opendnssec.org) which provides a PKCS#11 interface to a virtual HSM, implemented in the form of encrypted data on the local filesystem. It uses the Botan library for encryption and SQLite3 for data storage. Though less secure than a true HSM, it can provide more secure key storage than traditional key files, and can allow you to experiment with PKCS#11 when an HSM is not available.

The SoftHSM cryptographic store must be installed and initialized before using it with OpenSSL, and the SOFTHSM_CONF environment variable must always point to the SoftHSM configuration file:

$  cd softhsm-1.3.0 
$  configure --prefix=/opt/pkcs11/usr 
$  make 
$  make install 
$  export SOFTHSM_CONF=/opt/pkcs11/softhsm.conf 
$  echo "0:/opt/pkcs11/softhsm.db" > $SOFTHSM_CONF 
$  /opt/pkcs11/usr/bin/softhsm --init-token 0 --slot 0 --label softhsm 

SoftHSM can perform all cryptographic operations, but since it only uses your system CPU, there is no need to use it for anything but signing. Therefore, we choose the 'sign-only' flavor when building OpenSSL.

$ cd openssl-0.9.8s
$ ./Configure linux-x86_64 -pthread \
            --pk11-libname=/opt/pkcs11/usr/lib/libpkcs11.so \
            --pk11-flavor=sign-only \
            --prefix=/opt/pkcs11/usr

After configuring, run "make" and "make test".

Once you have built OpenSSL, run "apps/openssl engine pkcs11" to confirm that PKCS #11 support was compiled in correctly. The output should be one of the following lines, depending on the flavor selected:

        (pkcs11) PKCS #11 engine support (sign only)

Or:

        (pkcs11) PKCS #11 engine support (crypto accelerator)

Next, run "apps/openssl engine pkcs11 -t". This will attempt to initialize the PKCS #11 engine. If it is able to do so successfully, it will report “[ available ]”.

If the output is correct, run "make install" which will install the modified OpenSSL suite to /opt/pkcs11/usr.

Building BIND 9 with PKCS#11

When building BIND 9, the location of the custom-built OpenSSL library must be specified via configure.

Configuring BIND 9 for Linux with the AEP Keyper

To link with the PKCS #11 provider, threads must be enabled in the BIND 9 build.

The PKCS #11 library for the AEP Keyper is currently only available as a 32-bit binary. If we are building on a 64-bit host, we must force a 32-bit build by adding "-m32" to the CC options on the "configure" command line.

$ cd ../bind9
$ ./configure CC="gcc -m32" --enable-threads \
           --with-openssl=/opt/pkcs11/usr \
           --with-pkcs11=/opt/pkcs11/usr/lib/libpkcs11.so

Configuring BIND 9 for Solaris with the SCA 6000

To link with the PKCS #11 provider, threads must be enabled in the BIND 9 build.

$ cd ../bind9
$ ./configure CC="cc -xarch=amd64" --enable-threads \
            --with-openssl=/opt/pkcs11/usr \
            --with-pkcs11=/usr/lib/64/libpkcs11.so

(For a 32-bit build, omit CC="cc -xarch=amd64".)

If configure complains about OpenSSL not working, you may have a 32/64-bit architecture mismatch. Or, you may have incorrectly specified the path to OpenSSL (it should be the same as the --prefix argument to the OpenSSL Configure).

Configuring BIND 9 for SoftHSM

$ cd ../bind9
$ ./configure --enable-threads \
           --with-openssl=/opt/pkcs11/usr \
           --with-pkcs11=/opt/pkcs11/usr/lib/libpkcs11.so

After configuring, run "make", "make test" and "make install".

(Note: If "make test" fails in the "pkcs11" system test, you may have forgotten to set the SOFTHSM_CONF environment variable.)

PKCS #11 Tools

BIND 9 includes a minimal set of tools to operate the HSM, including pkcs11-keygen to generate a new key pair within the HSM, pkcs11-list to list objects currently available, and pkcs11-destroy to remove objects.

In UNIX/Linux builds, these tools are built only if BIND 9 is configured with the --with-pkcs11 option. (NOTE: If --with-pkcs11 is set to "yes", rather than to the path of the PKCS #11 provider, then the tools will be built but the provider will be left undefined. Use the -m option or the PKCS11_PROVIDER environment variable to specify the path to the provider.)

Using the HSM

First, we must set up the runtime environment so the OpenSSL and PKCS #11 libraries can be loaded:

$ export LD_LIBRARY_PATH=/opt/pkcs11/usr/lib:${LD_LIBRARY_PATH}

When operating an AEP Keyper, it is also necessary to specify the location of the "machine" file, which stores information about the Keyper for use by PKCS #11 provider library. If the machine file is in /opt/Keyper/PKCS11Provider/machine, use:

$ export KEYPER_LIBRARY_PATH=/opt/Keyper/PKCS11Provider

These environment variables must be set whenever running any tool that uses the HSM, including pkcs11-keygen, pkcs11-list, pkcs11-destroy, dnssec-keyfromlabel, dnssec-signzone, dnssec-keygen(which will use the HSM for random number generation), and named.

We can now create and use keys in the HSM. In this case, we will create a 2048 bit key and give it the label "sample-ksk":

$ pkcs11-keygen -b 2048 -l sample-ksk

To confirm that the key exists:

$ pkcs11-list
Enter PIN:
object[0]: handle 2147483658 class 3 label[8] 'sample-ksk' id[0]
object[1]: handle 2147483657 class 2 label[8] 'sample-ksk' id[0]

Before using this key to sign a zone, we must create a pair of BIND 9 key files. The "dnssec-keyfromlabel" utility does this. In this case, we will be using the HSM key "sample-ksk" as the key-signing key for "example.net":

$ dnssec-keyfromlabel -l sample-ksk -f KSK example.net

The resulting K*.key and K*.private files can now be used to sign the zone. Unlike normal K* files, which contain both public and private key data, these files will contain only the public key data, plus an identifier for the private key which remains stored within the HSM. The HSM handles signing with the private key.

If you wish to generate a second key in the HSM for use as a zone-signing key, follow the same procedure above, using a different keylabel, a smaller key size, and omitting "-f KSK" from the dnssec-keyfromlabel arguments:

$ pkcs11-keygen -b 1024 -l sample-zsk
$ dnssec-keyfromlabel -l sample-zsk example.net

Alternatively, you may prefer to generate a conventional on-disk key, using dnssec-keygen:

$ dnssec-keygen example.net

This provides less security than an HSM key, but since HSMs can be slow or cumbersome to use for security reasons, it may be more efficient to reserve HSM keys for use in the less frequent key-signing operation. The zone-signing key can be rolled more frequently, if you wish, to compensate for a reduction in key security.

Now you can sign the zone. (Note: If not using the -S option to dnssec-signzone, it will be necessary to add the contents of both K*.key files to the zone master file before signing it.)

$ dnssec-signzone -S example.net
Enter PIN:
Verifying the zone using the following algorithms:
NSEC3RSASHA1.
Zone signing complete:
Algorithm: NSEC3RSASHA1: ZSKs: 1, KSKs: 1 active, 0 revoked, 0 stand-by
example.net.signed

Specifying the engine on the command line

The OpenSSL engine can be specified in named and all of the BIND dnssec-* tools by using the "-E <engine>" command line option. If BIND 9 is built with the --with-pkcs11 option, this option defaults to "pkcs11". Specifying the engine will generally not be necessary unless for some reason you wish to use a different OpenSSL engine.

If you wish to disable use of the "pkcs11" engine — for troubleshooting purposes, or because the HSM is unavailable — set the engine to the empty string. For example:

$ dnssec-signzone -E '' -S example.net

This causes dnssec-signzone to run as if it were compiled without the --with-pkcs11 option.

Running named with automatic zone re-signing

If you want named to dynamically re-sign zones using HSM keys, and/or to to sign new records inserted via nsupdate, then named must have access to the HSM PIN. This can be accomplished by placing the PIN into the openssl.cnf file (in the above examples, /opt/pkcs11/usr/ssl/openssl.cnf).

The location of the openssl.cnf file can be overridden by setting the OPENSSL_CONF environment variable before running named.

Sample openssl.cnf:

        openssl_conf = openssl_def
        [ openssl_def ]
        engines = engine_section
        [ engine_section ]
        pkcs11 = pkcs11_section
        [ pkcs11_section ]
        PIN = <PLACE PIN HERE>

This will also allow the dnssec-* tools to access the HSM without PIN entry. (The pkcs11-* tools access the HSM directly, not via OpenSSL, so a PIN will still be required to use them.)

Warning

Placing the HSM's PIN in a text file in this manner may reduce the security advantage of using an HSM. Be sure this is what you want to do before configuring OpenSSL in this way.

IPv6 Support in BIND 9

BIND 9 fully supports all currently defined forms of IPv6 name to address and address to name lookups. It will also use IPv6 addresses to make queries when running on an IPv6 capable system.

For forward lookups, BIND 9 supports only AAAA records. RFC 3363 deprecated the use of A6 records, and client-side support for A6 records was accordingly removed from BIND 9. However, authoritative BIND 9 name servers still load zone files containing A6 records correctly, answer queries for A6 records, and accept zone transfer for a zone containing A6 records.

For IPv6 reverse lookups, BIND 9 supports the traditional "nibble" format used in the ip6.arpa domain, as well as the older, deprecated ip6.int domain. Older versions of BIND 9 supported the "binary label" (also known as "bitstring") format, but support of binary labels has been completely removed per RFC 3363. Many applications in BIND 9 do not understand the binary label format at all any more, and will return an error if given. In particular, an authoritative BIND 9 name server will not load a zone file containing binary labels.

For an overview of the format and structure of IPv6 addresses, see the section called “IPv6 addresses (AAAA)”.

Address Lookups Using AAAA Records

The IPv6 AAAA record is a parallel to the IPv4 A record, and, unlike the deprecated A6 record, specifies the entire IPv6 address in a single record. For example,

$ORIGIN example.com.
host            3600    IN      AAAA    2001:db8::1

Use of IPv4-in-IPv6 mapped addresses is not recommended. If a host has an IPv4 address, use an A record, not a AAAA, with ::ffff:192.168.42.1 as the address.

Address to Name Lookups Using Nibble Format

When looking up an address in nibble format, the address components are simply reversed, just as in IPv4, and ip6.arpa. is appended to the resulting name. For example, the following would provide reverse name lookup for a host with address 2001:db8::1.

$ORIGIN 0.0.0.0.0.0.0.0.8.b.d.0.1.0.0.2.ip6.arpa.
1.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0  14400   IN    PTR    (
                                    host.example.com. )